Initial commit of Descrybe v2 without local scratch artifacts.

Drop one-shot tmp/axe scripts and agent i18n scratch so the Gitea tree is deployable.
2026-08-09 22:47:43 +02:00
commit 8580c996c3
1285 changed files with 325780 additions and 0 deletions
@@ -0,0 +1,146 @@
+package processing
+
+import (
+	"context"
+	"os"
+	"testing"
+	"time"
+
+	"github.com/google/uuid"
+	"github.com/jackc/pgx/v5/pgxpool"
+)
+
+func TestCleanupExpiredSyncJobsKeepsLatestHashAndDeletesAged(t *testing.T) {
+	dsn := os.Getenv("DATABASE_URL")
+	if dsn == "" {
+		t.Skip("DATABASE_URL not set")
+	}
+
+	ctx, cancel := context.WithTimeout(context.Background(), 30*time.Second)
+	defer cancel()
+
+	pg, err := pgxpool.New(ctx, dsn)
+	if err != nil {
+		t.Fatal(err)
+	}
+	defer pg.Close()
+
+	companyID := uuid.New()
+	_, err = pg.Exec(ctx, `INSERT INTO companies (id, name) VALUES ($1, $2)`,
+		companyID, "retention-"+companyID.String()[:8])
+	if err != nil {
+		t.Fatalf("seed company: %v", err)
+	}
+	t.Cleanup(func() {
+		_, _ = pg.Exec(context.Background(), `DELETE FROM companies WHERE id = $1`, companyID)
+	})
+
+	insertFeed := func(name string) uuid.UUID {
+		t.Helper()
+		var id uuid.UUID
+		err := pg.QueryRow(ctx, `
+			INSERT INTO input_feeds (company_id, name, url, feed_type, status, sync_interval_minutes, options)
+			VALUES ($1, $2, '', 'csv', 'active', 60, '{}'::jsonb)
+			RETURNING id`, companyID, name).Scan(&id)
+		if err != nil {
+			t.Fatal(err)
+		}
+		return id
+	}
+
+	insertSync := func(feedID uuid.UUID, status, age, hash string) uuid.UUID {
+		t.Helper()
+		var id uuid.UUID
+		q := `
+			INSERT INTO feed_sync_jobs (
+				feed_id, company_id, status, started_at, completed_at,
+				updated_at, created_at, content_hash
+			) VALUES (
+				$1, $2, $3,
+				now() - interval '` + age + `',
+				now() - interval '` + age + `',
+				now() - interval '` + age + `',
+				now() - interval '` + age + `',
+				NULLIF($4, '')
+			) RETURNING id`
+		if err := pg.QueryRow(ctx, q, feedID, companyID, status, hash).Scan(&id); err != nil {
+			t.Fatal(err)
+		}
+		return id
+	}
+
+	assertGone := func(id uuid.UUID, label string) {
+		t.Helper()
+		var n int
+		if err := pg.QueryRow(ctx, `SELECT COUNT(*) FROM feed_sync_jobs WHERE id = $1`, id).Scan(&n); err != nil {
+			t.Fatal(err)
+		}
+		if n != 0 {
+			t.Fatalf("%s still present", label)
+		}
+	}
+	assertPresent := func(id uuid.UUID, label string) {
+		t.Helper()
+		var n int
+		if err := pg.QueryRow(ctx, `SELECT COUNT(*) FROM feed_sync_jobs WHERE id = $1`, id).Scan(&n); err != nil {
+			t.Fatal(err)
+		}
+		if n != 1 {
+			t.Fatalf("%s missing", label)
+		}
+	}
+
+	t.Run("deletesAgedWhenFresherHashExists", func(t *testing.T) {
+		feedID := insertFeed("retention-fresh")
+		olderHash := insertSync(feedID, "completed", "60 days", "hash-old")
+		midHash := insertSync(feedID, "completed", "45 days", "hash-mid")
+		oldFailed := insertSync(feedID, "failed", "45 days", "")
+		freshCompleted := insertSync(feedID, "completed", "1 day", "hash-fresh")
+		oldRunning := insertSync(feedID, "running", "45 days", "")
+
+		res, err := CleanupExpiredSyncJobs(ctx, pg)
+		if err != nil {
+			t.Fatal(err)
+		}
+		if res.SyncJobsDeleted < 3 {
+			t.Fatalf("sync_jobs_deleted=%d want >= 3", res.SyncJobsDeleted)
+		}
+
+		assertGone(olderHash, "older completed hash")
+		assertGone(midHash, "mid completed hash")
+		assertGone(oldFailed, "old failed")
+		assertPresent(freshCompleted, "fresh completed")
+		assertPresent(oldRunning, "old running")
+	})
+
+	t.Run("keepsNewestAgedHashWhenNoFresher", func(t *testing.T) {
+		feedID := insertFeed("retention-stale")
+		olderHash := insertSync(feedID, "completed", "60 days", "hash-old")
+		keepHash := insertSync(feedID, "completed", "45 days", "hash-keep")
+		oldFailed := insertSync(feedID, "failed", "45 days", "")
+
+		res, err := CleanupExpiredSyncJobs(ctx, pg)
+		if err != nil {
+			t.Fatal(err)
+		}
+		if res.SyncJobsDeleted < 2 {
+			t.Fatalf("sync_jobs_deleted=%d want >= 2", res.SyncJobsDeleted)
+		}
+
+		assertGone(olderHash, "older completed hash")
+		assertGone(oldFailed, "old failed")
+		assertPresent(keepHash, "newest aged completed with hash")
+
+		var hash string
+		err = pg.QueryRow(ctx, `
+			SELECT content_hash FROM feed_sync_jobs
+			WHERE feed_id = $1 AND status = 'completed' AND content_hash IS NOT NULL AND content_hash <> ''
+			ORDER BY completed_at DESC NULLS LAST LIMIT 1`, feedID).Scan(&hash)
+		if err != nil {
+			t.Fatal(err)
+		}
+		if hash != "hash-keep" {
+			t.Fatalf("lastContentHash=%q want hash-keep", hash)
+		}
+	})
+}